conversion to TT is a poor substitute to integral  open pen protection.

Both are flawed in my opinion - if in different ways. TT is entirely reliant on a delicate RCD operating reliably - which we know they don't always do.

On the other hand single phase open-PEN detection system that work by sensing the L-N/PE/PEN voltage straying outside normal limits are fundamentally flawed where a 3-phase distribution system is in use as there are a whole range of 3-phase not-quite-balanced balance situations where the PEN can be an awful way adrift from 0V but L-PEN voltage on a particular phase can still be within tolerance. Some manufacturers also add c.p.c. current monitoring (a bit like the old VoELCBs) to mitigate the danger a bit - but that relies on someone getting a shock before disconnection occurs. Then the whole thing is driven by electronics and a mechanism to release contacts ... very much like an RCD ...  so we're pretty much back to the same reliability question we have for TT. It's because of these flaws o-pen devices aren't recognised for use in installations with 3-phase available - fundamentally a much better job can be done by referencing to an artificial neutral point derived from all 3 phases.

If it's indoors, I'd stick with using the building's Earthing system and keep things simple.

   - Andy.

conversion to TT is a poor substitute to integral  open pen protection.

Both are flawed in my opinion - if in different ways. TT is entirely reliant on a delicate RCD operating reliably - which we know they don't always do.

On the other hand single phase open-PEN detection system that work by sensing the L-N/PE/PEN voltage straying outside normal limits are fundamentally flawed where a 3-phase distribution system is in use as there are a whole range of 3-phase not-quite-balanced balance situations where the PEN can be an awful way adrift from 0V but L-PEN voltage on a particular phase can still be within tolerance. Some manufacturers also add c.p.c. current monitoring (a bit like the old VoELCBs) to mitigate the danger a bit - but that relies on someone getting a shock before disconnection occurs. Then the whole thing is driven by electronics and a mechanism to release contacts ... very much like an RCD ...  so we're pretty much back to the same reliability question we have for TT. It's because of these flaws o-pen devices aren't recognised for use in installations with 3-phase available - fundamentally a much better job can be done by referencing to an artificial neutral point derived from all 3 phases.

If it's indoors, I'd stick with using the building's Earthing system and keep things simple.

   - Andy.

AJJewsbury said:

but that relies on someone getting a shock before disconnection occurs.

You mean, a bit like Additional Protection by RCD on Class II equipment or a live conductor in a severed flex? But in that case it's definitely the full line to Earth voltage, rather than "might be over 70 V"?

AJJewsbury said:

Then the whole thing is driven by electronics and a mechanism to release contacts ... very much like an RCD ...  so we're pretty much back to the same reliability question we have for TT. It's because of these flaws o-pen devices aren't recognised for use in installations with 3-phase available - fundamentally a much better job can be done by referencing to an artificial neutral point derived from all 3 phases.

Smoke and mirrors! I'd prefer to stick with 3 phases even if a car can cope with only one at a time.

AJJewsbury said:

On the other hand single phase open-PEN detection system that work by sensing the L-N/PE/PEN voltage straying outside normal limits are fundamentally flawed where a 3-phase distribution system

I've also been thinking about this statement. I think "fundamentally flawed" is far too strong - see below.

AJJewsbury said:

there are a whole range of 3-phase not-quite-balanced balance situations where the PEN can be an awful way adrift from 0V but L-PEN voltage on a particular phase can still be within tolerance.

But, with the voltage range in BS 7671, only on any one phase downstream of the broken PEN ... plus the fact that, in a real broken neutral situation, the neutral (PEN in this case) voltage moves around as loads change - we're definitely not talking about a static situation of perpetual non-detection.

So, I think the reality of the 722.411.4.1 (iv) device is perhaps Not ideal, but offers an improvement over doing nothing.